On primordial sense-antisense coding

Abstract

The genetic code is implemented by aminoacyl-tRNA synthetases (aaRS). These 20 enzymes are divided into two classes that, despite performing same functions, have nothing common in structure. The mystery of this striking partition of aaRSs might have been concealed in their sterically complementary modes of tRNA recognition that, as we have found recently, protect the tRNAs with complementary anticodons from confusion in translation. This finding implies that, in the beginning, life increased its coding repertoire by the pairs of complementary codons (rather than one-by-one) and used both complementary strands of genes as templates for translation. The class I and class II aaRSs may represent one of the most important examples of such primordial sense-antisense (SAS) coding (Rodin and Ohno, Orig Life Evol Biosph 25:565-589, 1995). In this report, we address the issue of SAS coding in a wider scope. We suggest a variety of advantages that such coding would have had in exploring a wider sequence space before translation became highly specific. In particular, we confirm that in Achlya klebsiana a single gene might have originally coded for an HSP70 chaperonin (class II aaRS homolog) and an NAD-specific GDH-like enzyme (class I aaRS homolog) via its sense and antisense strands. Thus, in contrast to the conclusions in Williams et al. (Mol Biol Evol 26:445-450, 2009), this could indeed be a "Rosetta stone" gene (Carter and Duax, Mol Cell 10:705-708, 2002) (eroded somewhat, though) for the SAS origin of the two aaRS classes.

Figure 1

The subcode for two sterically mirror modes of tRNA recognition by aminoacyl-tRNA synthetases. A: The condensed rearranged representation of the genetic code (Table 1), in which complementary codons are put vis-à-vis each other. The yin-yang-like pattern of the representation reveals the latent sub-code for the two modes of tRNA aminoacylation: (1) If the complementary codons contain YY vs. RR at the second and adjacent (either first or third) positions, their aaRSs recognize the tRNA acceptor from the same side of the groove, namely: minor (yellow) for 5’NAR3’ – 5’YUN3’ pairs, or major (blue) for 5’RGN3’ – 5’NCY3’ pairs; (2) If these positions are occupied by RY and YR, the modes of tRNA recognition are different, namely: minor (yellow) 5’YGN3’ vs. major (blue) 5’NCR3’ and major (blue) 5’NAY3’ vs. minor (yellow) 5’RUN3’. Precisely same rules are applicable to pairs of complementary anticodons. Taking into account the anticodon flanking 5’U and R3’ nucleotides allows us to show that in fact this sub-code minimizes a risk of confusion of tRNAs with complementary anticodons by aaRS, no matter are the latter real proteins or their putative ribozymic precursors. Other symbols: N and complementary и denote all four nucleotides; R, purine (G or A); Y, pyrimidine (C or U). For details, see (Rodin and Rodin, 2006b, 2008). B: The tRNA cloverleaf with complementary halves that are colored yellow (5’ half) and blue (3’ half), in accordance with the sub-code (A). Arrows show the two sides from which the putative ribozymic precursors of class I and class II p-aaRSs approached the proto-tRNAs. The 2^nd bases of triplets in the acceptor stem (marked red) and the anticodons show the concerted dual complementarity that may point to a common ancestor of the two codes they represent, the operational and classic ones, respectively (Rodin et al., 1996, 2009; Rodin and Rodin, 2006a). The 3’ strand of the acceptor arm represents the presumable ancestral palindrome self-templating and duplication of which readily form the extant tRNA cloverleaf (Rodin et al., 2009; see Rodin and Rodin, 2009 for details).

Figure 2

Inversion symmetry in the coding properties of the genetic code (after (Zull and Smith 1990). Codons for hydrophobic “core” amino acids are invariably antisense to polar amino acids found almost exclusively on protein surfaces. This arrangement is especially suited to producing molten globular gene products from both strands of a SAS gene. Note also that Proline and Glycine share one SAS codon-anticodon pair, such that a turn specified by a Pro-Gly sequence on one strand will be preserved on the opposite strand. The three groups represent the four patterns of pairs of complementary codons (and, symmetrically, anticodons) described by Rodin and Rodin (2006b) in their discussion of their relative ambiguity in a strand-symmetric RNA world.

Figure 3

The primordial SAS coding, gene duplications and complementarity of the diverged extant duplicates. A: The ancient in-frame SAS coding imposed strong constrains on evolution of complementary strand-genes (thick green and blue arrows). A duplication releases the daughter copies from these constraints. Their subsequent divergence with gradual silencing of the opposite strands (thin arrows) results eventually in two different genes (with sense strands, shown by thick arrows and antisense strand, shown by punctuated arrows) that may retain fingerprints of the original complementarity (on the right). B: The classic scheme of gene duplication without the primordial SAS coding: the diverged extant genes show no complementarity in anti-parallel “head-to-tail” alignment. C: The same as B but the original gene was a self-complementary palindrome. In this case, the extant offspring genes may also still display some complementarity derived from the original palindrome even though the latter had no SAS coding. However, the palindrome per se suggests possible descent from preceding SAS-encoded pairs of segments (shown in brackets by thick mini-arrows) that merged to constitute the gene duplicated later as a whole, i.e. the variant C is in fact a combination of the primordial SAS (A) and classic scheme of gene duplication (B).

Figure 4

Hsp70 vs. NAD-Gdh complementarity. Shown in the center (opposite directions) are two complementary sequences: the fragment of the HSP70 gene of the Achlya klebsiana and its (head-to-tail aligned) antisense complement, denoted AK-s and AK-as, respectively. Under the AK-s, the homologous sequence fragments of yeast (SSA2, SSA3 and SSB1) and bovine (Bos taurus, BT) HSP70 genes are aligned from left to right. Above the AK-as, from right to left, we mapped the homologous pieces of NAD-specific GDH genes of N. crassa (NC) (following the alignment in Fig. 2 from (Williams et al., 2009)) and S. cerevisiae (SC), as well as the 3-hydroxyacyl-CoA dehydrogenase (Type II HADH) of H. sapiens (HS). The putative (according to Williams et al., 2009) NAD-GDH homolog from the A.klebsiana’s close relative, the oomycete Aphanomyces euteiches (AE), is also shown, at the very top. Codons that share the same, complementary central nucleotide in the SAS alignment, according to the hypothesis of ancient SAS complementary coding, are colored green. Similarly marked are the cognate amino acids (some of them are identical). For the sense and anti-sense homologies of this A.klebsiana HSP70 region with two classes of aaRS, see (Carter and Duax, 2002).

Figure 4

Hsp70 vs. NAD-Gdh complementarity. Shown in the center (opposite directions) are two complementary sequences: the fragment of the HSP70 gene of the Achlya klebsiana and its (head-to-tail aligned) antisense complement, denoted AK-s and AK-as, respectively. Under the AK-s, the homologous sequence fragments of yeast (SSA2, SSA3 and SSB1) and bovine (Bos taurus, BT) HSP70 genes are aligned from left to right. Above the AK-as, from right to left, we mapped the homologous pieces of NAD-specific GDH genes of N. crassa (NC) (following the alignment in Fig. 2 from (Williams et al., 2009)) and S. cerevisiae (SC), as well as the 3-hydroxyacyl-CoA dehydrogenase (Type II HADH) of H. sapiens (HS). The putative (according to Williams et al., 2009) NAD-GDH homolog from the A.klebsiana’s close relative, the oomycete Aphanomyces euteiches (AE), is also shown, at the very top. Codons that share the same, complementary central nucleotide in the SAS alignment, according to the hypothesis of ancient SAS complementary coding, are colored green. Similarly marked are the cognate amino acids (some of them are identical). For the sense and anti-sense homologies of this A.klebsiana HSP70 region with two classes of aaRS, see (Carter and Duax, 2002).